Finance●●●Banger
Refund and Arbitration Protocol for Agents
Escrow refunds for failed AI API calls — Austin Griffith and Ethereum Foundation endorse it.
Zero to OneBig BrainSolve My Problem
AliAbdoli
202mo ago
Back to browse
GitHub Repository
Safe refund policy layer for AI agents — Python + TypeScript. Same behavior, shared tests.
2 starsPython
Refund Guard – a policy gate for AI agents that can issue refunds
by MJM_13·Apr 11, 2026·2 points·0 comments
AI Analysis
●●SolidSolve My ProblemShip It
MECE security map for refund agents when Stripe has no native guardrails.
Strengths
- •Agent input boundary restricts model to only amount and reason parameters
- •Provider invocation gate prevents refund calls unless policy passes 100%
- •Published on both PyPI and npm with matching test suites across languages
Weaknesses
- •Narrow scope—only refunds, not broader AI agent financial guardrails
- •Zero stars suggests early stage with unproven production adoption
Category
Target Audience
Developers building AI agents with payment capabilities
Similar To
Lakera · Guardrails AI · Rebuff
Similar Projects
AI/ML●●Solid
Governor – Quality gates for AI agents (465 tests, no Deps)
465 passing tests from 5,000+ production executions—guards that don't get tired.
Solve My ProblemDark Horse
june-jule
112mo ago
Security●●●Banger
AI-runtime-guard – Policy enforcement layer for MCP AI agents
Policy enforcement layer stops AI agents from deleting files or leaking credentials—no prompt retraining needed.
Solve My ProblemZero to One
JimmyRacheta
223mo ago
Security●●Solid
Fail-closed execution guard for AI agents (Python, pip installable)
Fail-closed execution guard with signed proofs—but risk scoring lacks published methodology or benchmarks.
Solve My Problem
echo_os
113mo ago
Security●●Solid
Agent Gate – Execution authority for AI agents, vault-backed rollback
The idea of inserting a deterministic 'gate' between proposed tool calls and execution is smart and practical: precomputed classification plus an agent-unreachable vault means destructive operations become reversible by default. The README calls out clear mechanics (envelope checks, vault backup, tiered responses and structured denials) which suggests this is more than a thought experiment — it's a focused infra piece for teams that actually let agents touch production.
Big BrainNiche Gem
hammer32
113mo ago
Developer Tools●●Solid
Agentpriv – Sudo for AI Agents
Minimal sudo wrapper for agents; solves real agent safety, but competes with framework guardrails.
Solve My ProblemShip It
nichkej
333mo ago